Molecular Rearrangements. IV. Triple-labeling Experiments on the

Molecular Rearrangements. IV. Triple-labeling Experiments on the Isotope Position Isomerization of 1,2,2-Triphenylethyl Acetate1. William A. Bonner, a...
0 downloads 0 Views 656KB Size
Jan. 5 , 1955

ISOTOPE POSITION ISOMERIZATION OF

solution was dried under vacuum, and then oxidized in 35 ml. of acetone, 10 ml. of water and 0.1 ml. of acetic acid with 750 mg. of KMnO4. After four days the mixture was worked up in the standard fashion. The benzophenone obtained was converted to its 2,4-DNPH (484 mg.) which was once crystallized from dioxane. The m.p. of the derivative was 237-238'. Radioactivity assay showed the 2,4-DN'I" to contain 0.0104 mc. per mole of carbon-14. Since the olefin Va, before treatment with formic acid yielded, after oxidation and derivative preparation, a 2,4DNPH of activity 0.00854 mc. per mole, it can be calculated that the olefin, on the above treatment, underwent only 0.72 yo isomerization. l-Pher~yl-C~~-P ,2-diphenylethyl p-Toluenesulfonate (IVc) -Thoroughly desiccated l-phenyl-l-CLL2,2-diphenylethmol (5.00 g.) was converted to its tosylate as previously described,* by allowing the sodium salt of the alcohol to react with p-toluenesulfonyl chloride in anhydrous ether. The crude product weighed 1.93 g., had m.p. 79-80", and was used without further purification. It has been shown previously2 that no isomerization attends the tosylation reaction so conducted. Acetolysis of l-Phenyl-C14-2,2-diphenylethyl+Toluenesulfonate (IVc) .-The tosylate IVc (1 .OO g., 2.33 mmoles) was added to a hot mixture of acetic acid (10 ml.) , acetic anhydride (0.1 ml.) and anhydrous sodium acetate (0.57 g., 6.98 mmoles, threefold excess). The mixture was heated for 45 minutes on a steam-bath, then diluted slightly with water and allowed to crystallize. The resulting 1,2,2-triphenyl-C1I4-ethyl acetate (0.61 g., 82%, m.p. 155-156') was deacetylated as before with lithium aluminum hydride which, after recrysproducing 1,2,2-triphenyl-C1I4-ethanol tallization from dilute acetic acid, weighed 0.46 g. (87%) and had m.p. 87-88'. The latter was oxidatively degraded in the usual manner to produce benzoic-phen~1-C~~ acid showing a m.p. of 122-123' and a radioactivity assay of 0.510 mc./mole after one recrystallization from water, and 2,4 dinitrophenylhydrazone benzophenone phenyl showing m.p. 240.5-241.5' and an assay of 0.4665 3z 0.003 mc per mole after two recrystallizations from dioxane. These values correspond to a 47.1 and 52.9% distribution of radioactivity, respectively. I t has been established previously2 that no further rearrangement of the acetate product occurs when sodium acetate is present in the reaction mixture to buffer the p-toluenesulfonic acid produced during acetolysis. Hydrolysis of l-Phenyl-CI4-2,2-diphenylethyl $-Toluenesulfonate.-The ring-labeled tosylate IVc (0.93 9.) was dissolved in an acetone (25 m1.)-water ( 8 ml.) mixture, which was then refluxed for 2.5 hours. After allowing the mixture t o stand overnight the acetone was evaporated in an airsteam, water was added and the product was extracted into ether. The ethereal solution was washed with water, dried

.

-

-

-

[CONTRIBUTION FROM

THE

1,2,2-TRIPHENYLETHYLACETATE

99

and the solvent evaporated, yielding 0.60 g. (100%) of crude 1,2,2-triphenyl-C114-ethanol.This was recrystallized The from dilute acetic acid to give 0.58 g., m.p. 87-88'. purified carbinol was oxidized with permanganate to give benzoic-phenyl-C14 acid, m.p. 121-122', assay 0.728 mc. per mole after one recrystallization, and benzophenonephenyl-Cl'4 2,4-dinitrophenylhydrazone, m .p. 240.5241.5', assay 0.2330 and 0.2340 (average 0.2335 f 0.0005) after two recrystallizations. These values correspond to a radioactivity distribution of 76.5 and 23.6%, respectively. Action of p-Toluenesulfonic Acid on 1,2,2-Triphenylethsee if the p-toluenean01-l-C14 (IIc) in Acetone-Water.-To sulfonic acid produced during hydrolysis of tosylate IVc might have any further isomerizing action on the carbinol (0.60 g., radioactivity formed, 1,2,2-triphenylethanol-l-C1* assay 1.000 mc. per mole) was dissolved for 2.5 hours in a refluxing mixture of acetone (25 ml.) , water (8 ml .) and p toluenesulfonic acid hydrate (0.42 g., 1 molar equivalent). The carbinol was re-isolated and its oxidation was conducted as described above. The benzophenone derivative from the oxidation showed a radioactivity assay of 0.0119 mc. per mole. Thus the isomerized carbinol produced on tosylate hydrolysis undergoes no further isomerization under the reaction conditions. Dehydration of l,Z,Z-Triphenylethanol (IIa) and (IIc) in Xylene with p-Toluenesulfonic Acid.-Ring-labeled 1,2,2triphenylethanol (IIc) (700 mg.) was dissolved in 27 ml. of dry xylene containing 550 mg. of anhydrous p-toluenesulfonic acid. This mixture was boiled one hour, at which time excess water was added to the flask, and the separated xylene layer was diluted with ether, washed with bicarbonate solution and concentrated. Hexane was added to the concentrate, and the solution was passed through a column of alumina, and the eluate was evaporated yielding 312 mg. of olefin Vcd, m.p. 68-70'. No depression with authentic sample. The benzophenone 2,4dinitrophenylhydrazone obtained upon oxidation of the 312 mg. of olefin Vcd, followed by preparation of the derivative was assayed for carbon-14 with the following results: 0.5975, 0.5965 mc. per mole, average 0.5970 f 0.0005 mc. per mole, corresponding to 60.2% of the original radioactivity of IIc. The benzoic acid fraction was discarded. A reaction similar in every respect was carried out on the chain-labeled 1,2,2-triphenylethan01-1-C~~ (IIa) The benzophenone 2,4dinitrophenylhydrazone resulting was assayed for carbon-14 with the following results: 0.4772, 0.4778, 0.4815, average 0.4788 0.0017 mc. per mole, corresponding to 47.88% of the original radioactivity. The benzoic acid fraction, assayed as a rough check on the total radioactivity, contained 0.5370 mc. per mole (single determination). OAK RIDGE,TENNESSEE

.

CHEMISTRY DIVISIONOF OAK RIDGENATIONAL LABORATORY]

Molecular Rearrangements. IV. Triple-labeling Experiments on the Is0tope Position Isomerization of 1,Z,Z-Triphenylethyl Acetate' BY WILLIAMA. BONNERAND CLAIRJ. COLLINS RECEIVED APRIL26, 1954 1,2,2-Triphenylethyl acetate, separately labeled in the chain, phenyl and acetate portions of the molecule, has been studied kinetically with respect to the rates of radiochemical isomerization and of acetoxyl exchange under the influence of an acid catalyst. Confirmatory experiments, using combinations of these three differently labeled species, have also been carried out. It is found that the rates of radiochemical equilibration of the ring-labeled and chain-labeled acetates are equal, and each is identical to the rate of loss of the labeled acetoxyl group. The data are explainable in terms of open carbonium ion intermediates in which the cation undergoes radiochemical isomerization prior to product formation. I t is shown that, as expected, no internal return accompanies the isomerization.

Introduction In the previous paper2 a study was reported of a number of irreversible carbonium ion-type proc(1) This paper is based upon work performed under Contract Number W-7405-eng-26 for the Atomic Energy Commission at Oak Ridge National Laboratory. Reprint requests should be addressed to C. J. C. (2) C. J. Collins and W. A. Bonaer, THIS JOURNAL, 7 7 , 92 (1955).

esses with doubly labeled 192,2-triphenylethyl derivatives. That is, a comparison was made of the distributions of the radioactive labels when derivatives of lf2,2-triphenylethanol-1-C and of l-phenyl-C14-2,2-diphenylethanol undergo certain irreversible carbonium ion-type reactions. These reactions have included tosylate acetolysis, tosylate

100

WILLIAMA. BONNERAND CLAIRJ. COLLINS

hydrolysis, carbinol dehydration and elimination of acetic acid from the acetates. The observed radiochemical results could be explained completely and quantitatively in terms of open carbonium ions which undergo radiochemical isomerization to a varying degree depending on their lifetimes (eq. 1). There was no evidence for the existence of bridged in any of the reactions studied. In

Vol. 77

further information might be obtained regarding the validity of the proposed mechanism.2 T o this end, the triply labeled system of 1,2,2-triphenylethyl acetates, Ph&HC*H(OAc)Ph (I, chainlabeled), Ph&HCH(OAc)Ph* (11, ring-labeled) and Ph2CHCH(OCOC*H3)Ph(111, acetyl-labeled), has been studied kinetically with regard to the comparative rates of label migration and of acetoxyl exchange. Xethyl rather than carbonyl ca 63 Ph-CH-C*H-Ph* Ph-CH-C*H-Ph* labeling in the acetoxyl group of I11 was selected 1 I in order to reduce any possible complications which Ph Ph might arise from an isotope effect during acetoxyl * 63 removal. Ph-CH-C*HPh (1) II Methods and Results Ph The kinetic studies were conducted a t 54.6 + certain reactions, e.g., formic acid catalyzed de- 0.4", using the appropriate 1,2,2-triphenylethyl hydrations of chain- and ring-labeled2 l ,2,2-tri- acetate, in acetic acid solution containing one phenylethanols, the intermediates seemed to be equivalent of p-toluenesulfonic acid. Aliquots of sufficiently long-lived to permit statistical redistri- the reaction mixture were then quenched in aqueous bution of the radioactive labels prior to irreversible sodium chloride solutions after various time information of the final olefinic products. In other tervals. The acetate from each aliquot was rereactions, for example, hydrolysis or acetolysis of covered and, for those runs in which i t was neceschain- and ring-labeled 1,2,2-triphenylethyl p sary to determine carbon-14 a t this point, crystaltoluenesulfonates, the shorter-lived intermediates lized one or two times. The crude acetate obwere permitted only a partial equilibration before tained quantitatively on quenching the aliquots occurrence of the final irreversible step. In such was quite pure; recrystallization prior to radioreactions, the chain-labeled compouqd proceeded activity assay was carried out chiefly to remove p faster toward equilibrium than did the ring-labeled toluenesulfonic acid and sodium chloride which compound and the degree of rearrangement ob- might possibly have been occluded during the served during reaction in the chain-labeled system quenching procedure. The extent of carbon-14 could be calculated with considerable precision, equilibration in the chain- and ring-labeled meknowing the extent of rearrangement occurring during reaction in the phenyl-labeled system. The tates was ascertained by deacetylating the recovered acetate, oxidizing the resulting carbinol calculations were based on a simple kinetic relation- and assaying the benzophenone oxidation product ship involving equilibrating cationic intermediates. for radioactivity. The extent of acetoxyl exchange Acetoxyl group removal from 1,2,2-triphenyl- of the acetyl-labeled acetate with its non-radioethyl acetate appears to be a process in which the active environment was ascertained by assaying equilibration of labeled phenyl groups is complete, for total radioactivity the acetate recovered from during the lifetime of the intermediates.2 Thus each aliquot. Both isomerization reactions, as when l-phenyl-C14-2,2-triphenylethyl acetate (11) well as the acetoxyl exchange reaction, were found is heated with formic acid with or without 9-tolu- to obey first-order kinetic expressions with respect enesulfonic acid as catalyst, the triphenylethylene to acetates I, I1 or I11 under the reaction conditions obtained is found2to have a statistical distribution employed. First-order rate constants a t the time of of the radioactive label each aliquot removal were calculated as described in the Experimental section. The results of these kinetic experiments are presented in Tables 1-111. 2/3 1/3 Examination of Tables 1-111 indicates that the In addition, both 1,2,2-triphenylethy1-1-Cl4 ace- first-order rates of equilibration of the chaintate (I) and 1-phenyl-C*4-2,2-triphenylethyl acetate labeled acetate and the ring-labeled acetate are (11), when warmed in acetic acid containing p- equal within experimental error and each is essentoluenesulfonic acid, give 1,2,2-triphenylethyl ace- tially the same as the first-order rate of acetoxyl tate samples having a statistical distribution of exchange for the acetyl-labeled acetate, although radioactivity.2 The isomerization of this acetate the data for acetoxyl exchange seem to be subject labeled in various positions seemed to offer promise to a larger external error than do the data of Tables in the study of internal return in the present I or 11. These results demonstrate that during the system for at least one set of conditions. By a p-toluenesulfonic acid catalyzed isomerization of comparison of the rates of radioactivity redistribu- labeled 1,2,2-triphenylethyl acetates, statistical tion both of the chain-labeled and of the ring- equilibration of each label occurs each time a n acetlabeled acetates with the rate of acetoxyl exchange oxyl group i s removed from the discretely labeled with environment during this reaction, it was starting material. hoped that the presence or absence of internal Although the possibility was considered exreturn could be demonstrated, and that, in addition, tremely unlikely, there was a chance that a con(3) D. J. Cram, THISJOURNAL, 71, 3863 (1949). sistent external error was operative during the (4) J. D.Roberts and C.M. Regan, ibid., 76, 2069 (1953). gathering of the data of Tables I and I1 which (5) (a) S. Winstein and D. Trifan, ibid., 74, 1154 (1962); (b) D.J. would make it seem that the rates of equilibration Cram and F. A. Abd Elhafez, i b i d . , 7 6 , 3189 (1953). of the ring-labeled and chain-labeled acetates (6) F. A. Abd Elhafez and D. J. Cram, ibid., 76, 339 (1953).

Jan. 5, 1955

ISOTOPE POSITION ISOMERIZATION O F

1,2,2-TRIPHENYLETHYLACETATE

101

Discussion

TABLE I CONSTANTS FOR

The kinetic data of the preceding section allow the following conclusions to be drawn concerning 1,2,2-TRIPHENYLETHYL-l-C'* ACETATE kr X lo*, hours-1 the nature of the intermediates in the isomerization of the several labeled species of 1,2,2-triphenylRun 1 2.22 f 0 . 0 2 ethyl acetate: (1) Internal return6 does not occur 2.22 f .06 Run 2 during the isomerization, since the rate of acetoxyl Average 2.22 f .05 loss is the same as the rate of equilibration of the TABLE I1 acetates I and 11. If internal return should take place during these reactions, then the rate of acetFIRST-ORDER RATECONSTANTS FOR THE EQUILIBRATION OF 1-PHENYI.-C'4-2,2-DIPHENYLETHYL ACETATE oxy1 loss, KIII, would be expected to be considerably k i i X 102, hours-1 slower than the two equilibration rates kI and ~ I I . 2.22 f 0.09 This conclusion might have been anticipated, since Run 1 2.08 f .07 Run 2 the leaving group is undoubtedly an acetic acid molecule. Thus an ion pair is not formed on ionizaAverage 2.15 f .08 tion and ion-pair collapse is, of course, impossible. TABLE 111 These same considerations do not apply to the FIRST-ORDER RATE CONSTAXTS FOR ACETOXYLEXCHANGEsolvolyses reported in the preceding paper, during O F 1,2,2-TRIPHENYLETHYLACETATE-2-C'4 which i t is possible that internal return does occur. kIII X 103, hours-' (2) Symmetrical3 or unsymrnetrica14~6bridged Run 1 1.97 k 0.06 ions can be ruled out as the sole cationic interRun 2" 2.56 .03 mediates since, in the absence of internal return, Run 3' 2.39 f .07 and whether or not they demonstrate stereo~pecificity,~ their postulation would demand that k I Average 2 31 f .05 should be greater than kII.* In the event that the Value obtained running the ring-labeled and acetyllabeled 1,2,2-triphenylethyl acetates in the same reaction trans-phenonium ion were formed to the complete vessel. * One-half of the value obtained using two equiva- exclusion2 of the cis-phenonium ion, then the ringlents of p-toluenesulfonic acid. labeled acetate I1 could never exhibit more than were equal when in reality they should bear the 50% rearrangement of the radioactive label. (3) A completely concerted process in which one 0.75 ratio which might be predicted on the basis of the previously derived kinetic reaction.2 This phenyl group migrates totally as it assists in the possibility has now been definitely and completely removal of an acetoxyl group excluded by the results of an experiment in which H OH I P I the two differently labeled acetate samples, I and Ac-0 ( 0-C-C*Hs 11, were mixed homogeneously in solution and then L I @ subjected to the conditions of the acid-catalyzed Ph-CH-C*H-Ph* I 7 isomerization in the same reaction vessel. The Ph' degradative procedure and the radiochemical results of this experiment are outlined in Chart I. is excluded as an explanation for the present data Radioactivity data are shown for each molecule or since it would require that the rate of isomerization for each portion thereof as millicuries per mole. of both the ring-labeled and chain-labeled acetates The reaction conditions were identical to those be equal to the rate of acetoxyl exchange (KIII). employed in gathering the data of Tables I and I1 Such a mechanism would then require that the except that only one aliquot was removed-after rate of equilibration ( K I ) of the chain-labeled acetate 72 hours. From these data KI is calculated to be be twice the rate of acetoxyl exchange, and the 0.0203 f. 0.0012 hour-l, while KII is calculated as rate of equilibrium ( K I I ) for the ring-labeled acetate 0.0211 f 0.008 hour-'. The errors in the values be 3/2 the rate of acetoxyl exchange, again conof the two constants are calculated from the trary to observation. average errors of the radioactivity determinations (4)The general mechanism proposed in the preshown in Chart I. Allowing for these average vious paper2 in which the classical carbonium ion errors, the ratio k I I / k I = 1.04 lies between the intermediates undergo more or less equilibration, extreme limits 0.945 and 1.14, excluding the ratio depending upon their lifetimes, receives support. 0.75 as a possibility. During the isomerizations of acetates I and 11, it is obvious from the present kinetic data, that the CHARTI OAc 1 OAc ionic intermediates are sufficiently long-lived in

FIRST-ORDER R A T E

THE

EQUILIBRATION OF

0

I

/

PhZCHC *H/Ph* 0.5130 f 0.0001\0.4825 f 0.0001

p-TSA

___f

HOAc

1

$h26HCHPi I

* *

PhCOOH

+

PiNH2 P*hC;OH 0.3227 f 0.0010

H2SO4

x

*

J

t--Ph,C=NOH

+ hh2&=0

I

7

2,4-DNPH 0.4484 i 0.0011

(7) P. I. Pollak and D. Y. Curtin, THISJOURNAL, '72, 961 (1960); D. Y.Curtin and P. I. Pollak, ibid., 73, 992 (1951); D. Y. Curtin, E. E. Harris and P. I. Pollak, ibid., '73, 3453 (1961); D.Y.Curtin and E.K. Meislich, ibid., '74, 6906 (1962); D. Y. Curtin and D. B. Kellom, ibid., 75, 6011 (1953): D. J. Cram and F. A. Abd Elhafez, ibid., '76, 80 (1954). See also discussion of stereospecificity in ref. 2. (8) A symmetrical phenonium-ion mechanism in which a complete lack of stereospecificity is postulated would require that the rate of equilibration for the chain-labeled acetate be given by the relation kr = 2k = (l/;) In ( x , / x e x ) , and t h a t for the ring-labeled acetate y) where k, ze, x , ye by the relation of kII = 3k/2 = ( l / t ) In ye/& and y have the same significance as in the identical relations derived in the preceding (ref. 2) paper. Thus, k~:krrshould be in the ratio 4:3.

-

-

102

WILLIAMA. BONNER AND CLAIRJ. COLLINS

acetic acid solution, in the absence of acetate ion, to permit statistical redistribution of both the chain and ring labels before reaction of these intermediates with non-radioactive acetic acid. Thus, each mole of acetate product is fully equilibrated radiochemically for each mole of radioactive acetoxyl lost.

Experimental synthesis for this acetate was employed which differs slightly from that pre(4.5 9.) viously reported.9 1,2,2-Triphenylethan01-1-C~~ was dissolved in pyridine (20 ml.) and the solution was treated with acetic anhydride (15 ml.). The mixture was allowed to stand for 24 hours, then diluted with water and the product was filtered, rinsed and dried. The crude product was dissolved in excess hot acetone, then the solution was concentrated and treated with an equal volume of hot ethanol. The pure acetate (5.14 g., 99%, m.p. 155-156') crystallized as splendid needles. In other experiments giving comparable results, the original reaction mixture was allowed t o stand for one hour, then heated on the steambath for one hour prior t o dilution. Such procedures were employed in the synthesis of all three alternatively labeled acetates studies except, of course, that radioactive acetic anhydride was employed in the preparation of the acetyllabeled acetate. The radiochemical structure of one of the acetates prepared by this method was confirmed by deacetylation and oxidative degradation of the resulting carbinol. Isolation and assay of the benzophenone fraction from the oxidation by the usual procedures indicated that only 0.66% of the total radioactivity was found in the supposedly non-radioactive fraction, a result in accord with our previous observations.

Vol. 77

where k~ is the specific rate for the conversion of 1,2,2-triI into an equilibrium mixture of 1,2,2~ h e n y 1 - l - Cacetate ~~ triphenylethyl-1 ,2-C1I4 acetate, x is the radioactivity assay of the benzophenone fraction isolated a t time t , and x, is the radioactivity assay of the benzophenone fraction a t complete equilibrium (0.500 timcs the assay of the starting acetate). The rates of equilibration for the ring-labeled acetate werc calculated by means of the equation

1,2,2-Triphenylethyl-1-Cl4 Acetate.-A

k r r = 2.303- log

Ye - Y where krr is the specific rate constant for the conversion of acetate I1 t o its equilibrium mixture, y is the radioactivity assay of the benzophenone fraction at time 1, and ye is the assay of the benzophenone fraction a t complete equilibrium (0.667 times the assay of the starting acetate). The rates of acetoxyl exchange for the acetyl-labeled acetate were calculated by the first-order expression

2.303 log -5.a--x t where a is the original radioactivity assay of the starting acetyl-labeled acetate, and x is the quantity of radioactive acetoxyl lost at time t . The radioactivity assay of the acetate recovered a t time t thus defines the value of a - x . Kinetic Data.-In Tables IV-1.1 are given the equilibration or exchange rates calculated for each time t from the corresponding radioactivity assays of the products isolated a t each time t. The averages of these values appear in Tables 1-111. kIII =

~

TABLE IV EQUILIBRATION CONSTANTS FOR 1.2,2-TRIPHEXYLETHYL-1RADIOACTIVITY ASSAYS OF THE THREEALTERNATIVELY CI4 ACETATE Assay of SPECIES LABELED ACETATE Assay, mc./mole

1,2,2-Triphenylethy1-l-Cl4 acetate 2.161 rt 0.006 l-Pher1yl-C~~-2,2-diphenylethyl acetate 0,991 i ,004 1,2,2-Triphenylethyl acetate-2-C14 0.549 ,003

*

Kinetic Investigations.-The kinetic experiments leading to the rate data in Table 1-111 were conducted as follows. The 1,2,2-triphenylethyl acetate under study (3.000 9.) was dissolved in hot glacial acetic acid (70.00 ml.) and the solution was allowed t o reach the thermostat temperature of 54.6 i 0.4". At this point 6.00 ml. of catalyst solution was added, and the time was recorded. The catalyst solution consisted of 9.89 g. of desiccated p-toluenesulfonic acid dissolved in 32.4 ml. of glacial acetic acid containing 4.00 nil. of acetic anhydride. The reaction mixture was thus 0.1260 M in p-toluenesulfonic acid and 0.1250 M in starting acetate. At various time intervals 15.0-ml. aliquots were pipetted from the thermostated reaction mixture and drained into water. I n those runs in which the recovered acetate was assayed for carbon-14, the precipitate was filtered and crystallized from an acetone-alcohol mixture. For acetoxyl-exchange data the acetate was assayed a t this point, or after a second recrystallization. For phenyl equilibration data the acetate (including that recovered from the recrystallization mother liquors) was deacetylated with ethereal lithium aluminum hydride, and the crude carbinol obtained was oxidized in the usual fashion2 with an acetonewater solution of potassium permanganate. The benzophenone oxidation fraction was isolated as usual and converted t o its 2,4-dinitrophenylhydrazone,which was recrystallized once or twice from dioxane prior t o radioactivity assay. In one acetoxyl exchange run the volume of catalyst solution employed was doubled and the volume of acetic acid decreased accordingly, giving a solution 0.2714 M in ptoluenesulfonic acid and 0.1250 M in starting acetate. In several of the experiments the size of the run was scaled to exactly 4 / 3 that indicated above. Kinetic Calculations.-The rates of equilibration for the chain-labeled acetate were calculated by means of the integrated first-order expression (9) W. A. Bonuer and C. J. Collins, THISJOURNAL, 76, 5376 (1953).

Run

1

starting acetate, mc./mole

2.161 ( x , 1.081)

Time, hours

PhtC*O assay ( x ) ,

kr X 102,

mc./mole

G

0.172

22 30 46 70

.529 ,689 ,848

hours - 1

.419

.

(2.900) 2.229 2.243 2.207 2.192

~ ~

~

-~.

2.218f0.018 2

1.000 (x, 0.5001

9 21 44 48 52 57.5 70 iA 95

0.099A .191 ,314

,317 ,347 ,365 ,398 .402 .435

(2.460) 2.296 2 ' 250 2.130 2.280 2.272 2.2fi8

2 . 13f) 2.164

2.224 f 0 062 OAc I

Isomerization of Ph*CHC*HPH*.-The doubly labeled acetate (2.4459 g. containing 0.5130 2~ 0.001 mc./mole carbon-14 in the C-1 ethyl carbon and 0.4825 f 0.0001 mc./mole in the C-1 phenyl group) was dissolved in 57 ml. of glacial acetic acid in a 200-ml. flask and. after having been placed in a thermostat a t 54.6 & 0.4" for one hour, 4.9 ml. of the catalyst solution (described previously in this section) was added, and the mixture was well shaken t o ensure homogeneity. After 72 hours, a 32-ml. aliquot was removed, and poured into excess water containing sodium chloride t o congeal the precipitate. The acetate was filtered, dried and added t o an ethereal solution of lithium aluminum hydride, from which the carbinol was recovered. The crude carbinol was dissolved in 60 ml. of acetone, anti to it was added 3.8 g. of KMn04 in 32 ml. of water containing 0.2 ml. of acetic acid. After four days, the oxidation mixture was worked up in the usual fashion. The benzo-

.

1-HEXYNEWITH SOME ALKYLAND ALKENYLMAGNESIUM HALIDES

Jan. 5, 1955

TABLE V CONSTANTS FOR

EQUlLIBRATION

DIPHBNYLETHYL

Run

1

Assav of starling acetate, mc./mole

I-PHENYL-C1'-3,2-

ACETATE

TABLE VI ACETOXYLEXCHANGE CONSTANTS FOR 1,2,%TRIPHENYL ETHYL ACETATE-2-C1'

1.001)x

Time, hours

0.991 6 ( ~ ~ 0 . 6 6 1 ) 22 30 46 71.5

Phr*CO assay (Y), mc./mole

k1I

Assay of recovered acetate

x IO',

hours -1

0.107 .270 . 318 ,419 ,510

(2.950) 2.385 2.190 2.186 2.100

-

Run

Assay of starting acetate, mc./mole

Time (a x) houri mc./mole

krrx X 103,

1

0.549

6 . 5 0.483 23 ,344 30.75 .293 47 .224 54 .198

1.982 2.035 2.042 1.905 1.890

2

0.495

6 22 30 46 71.5

0.282 .278 ,230 ,153

0.549 4.5 (p-toluenesulfonic acid 23 concn. doubled) 27 30

0.438 .178 .151 .137

2.215f0.092 2

0.991 9 ( ~ ~ 0 . 6 6 1 ) 21 44 48 52 57.5 70 76 95

2.208 2.162 2.102 2.120 2.062 2.045 2.028 2.018 1.978

0.119 .241 .399 .414 .435 .457 ,501 .518 .,560

2.080=t00.070 phenone fraction was dissolved in 8 ml. of ethanol treated with Norit and filtered through Celite, then 1.5 mf. of this solution was converted to the 2,4-dinitrophenylhydrazone (189 mg.) which was twice crystallized from dioxane and dried in an Abderhalden. The radioactivity assay was: 0.4495; 0.4473 average 0.4484 f 0.0011 mc./mole. The remaining 6.5 ml. of ethanolic benzophenone was heated under reflux for 24 hours with 1 ml. of pyridine and 0.5 g. of hydroxylamine hydrochloride. The mixture was then poured into water and filtered, yielding 647 mg. of crude benzophenone oxime. This was not purified, since several model experiments had demonstrated that the crude oxime gave satisfactory yields in the Beckmann rearrangement. The oxime was dissolved in 12 ml. of concd. sulfuric acid and the solution was warmed on a steam-bath for one hour, after which time the cooled solution was poured over crushed ice. The resulting mixture was extracted three times with ether, the ether was clarified and concentrated, yielding 232 mg. of pure benzoic acid. This was twice crystallized from water, and identified by melting point and mixed

[CONTRIBUTION NO. 932 FROM

THE

103

hours-]

1.9711t 0.059 (4.318) 2.621 2.560 2.553 2.500

.083

2.558 f 0.032 3

(Divided by 2)

5.00 4.90 4.77 4.51 4.79 f 0 . 1 5 2.39 rt 0.07

melting point. The radioactivity assay was: 0.3238; 0.3217, average 0.3227 f 0.0010 mc./mole of carbon-14. From these data it was calculated that k1 = 2.03 =t 0.12 X lo-* hours-', and k11 = 2.11 f 0.08 X hours-', and that k I I / k l = 1.04 =k 0.095. Radioactivity Determinations.-These were carried out in the normal fashion as described in the previous paper.* For the determination of the isomerization of doubly labeled acetate (I and 11) described in the preceding section, the highest precision was necessary. This was obtained by weighing large samples (20-50 mg.) of analytical material on a micro-balance, and using the technique described for previous double-labeling experiments.l OAKRIDGE, TENNESSEE

DEPARTMENT OF CHEMISTRY, UNIVERSITY

OF PITTSBURGH]

The Reaction of 1-Alkynes with Organometallic Compounds. I. Some Alkyl and Alkenyhagnesium Halides

1-Hexyne with

BY JOHN H. WOTIZ,C. A. HOLLINCSWORTH AND RAYMOND DESSY' RECEIVED MARCH17, 1954

In the reactions of some Grignard reagents with 1-alkynes the rate of the evolution of gas was measured. The following relative rates were found: MeMgI, 6; MeMgBr, 6 ; MeMgC1, 16; n-PrMgBr, 59; EtMgI, 71; EtMgBr, 100; EtMgC1, 155; i-PrMgBr, 210; AllylMgBr, 435. A possible explanation of these relative rates is proposed.

In our laboratory we had many times the opportunity to prepare 1-hexynylmagnesium bromide in five mole quantities. The preparation consists of treating 1-hexyne (I) with a Grignard reagent such as methyl (11) or ethylmagnesium bromide (111) in boiling ether. The completeness of reaction was noted by the termination of evolved gas (methane C'HgC=C-H

+ R-MgX R

----f

CpHgC=C-MgX CHI (11), CzHs (111)

+ R-H(gas)

(1) National Science Foundation Predoctoral Fellow.

or ethane). We have observed that the reaction with ethylmagnesium bromide was finished shortly after the completion of addition whereas with methylmagnesium bromide we needed more than 24 hours before we considered the reaction to be finished. It was the purpose of this investigation to establish the rate of reaction of some Grignard reagents with l-alkynes. The literature contains two studies of the relative reactivities of Grignard reagents with compounds containing an "active" (acidic) hydrogen. Thus, the reactions with chloromagnesium phenyl-